Subjective forecasts of the Industrial Production Index based on the confidence of the manufacturing companies

Summary of Background Data The Industrial Production Index is one of the most important indicators to evaluate the state of health of the economy of a country. Dealing with official data, the issue of the delay in the publication of the provisional estimate is a very actual theme. For this purpose, it appears fundamental to have anticipations of these indicators. Many contributions have tried to give predictions about the future tendency of this quantity using different statistical models based on possible relationships with other micro or macro-economic variables. This study proposes a forecast approach based on the subjective expectations on the future given by the entrepreneurs of manufacturing companies. Objectives The main aim of this contribution is to give a systematic approach to obtain very-short term predictions for the Industrial Production Index using the long-term relationship between this time- series and the confidence of the manufacturing companies in terms of judgement and expectations. Methods From a methodological point of view, an approach for multivariate time-series is proposed, searching for a co-integration link between the two considered time-series. Once obtained this long- term relationship, the latter is used in combination with the expectations of the entrepreneurs to obtain the short-term forecasts. Results The proposed approach has been applied on the two time-series from 2010 to 2024. The preliminary results show a similarity between judgment and expectations of the manufacturing companies about the production. This is a mandatory condition to use the expectations as a predictor of the Industrial Production Index. The short-term forecasts obtained are in line with the expected values of the entire time-series. Discussion/Conclusions The proposed approach has been used to detect short-term movements for the Industrial Production Index in order to anticipate possible outliers in the time series due to possible crisis.